alex-hse-repository

HORIZON = 62
pipeline = Pipeline(
    model=CatBoostModelPerSegment(),
    transforms=[
        BinsegTrendTransform(in_column="target", n_bkps=2, min_size=150),
        DateFlagsTransform(day_number_in_week=True, day_number_in_month=False, is_weekend=False),
    ],
    horizon=HORIZON,
)

alex-hse-repository

HORIZON = 62
pipeline = Pipeline(
    model=CatBoostModelPerSegment(),
    transforms=[
        LinearTrendTransform(in_column="target"),
        DateFlagsTransform(day_number_in_week=True, day_number_in_month=False, is_weekend=False),
    ],
    horizon=HORIZON,
)

alex-hse-repository

pipeline = Pipeline(
    model=CatBoostModelPerSegment(),
    transforms=[
        DateFlagsTransform(day_number_in_week=True, day_number_in_month=False, is_weekend=False),
    ],
    horizon=HORIZON,
)

metrics, forecast, _ = pipeline.backtest(ts, metrics=[SMAPE()], n_folds=3)
plot_backtest(forecast, ts, history_len="all")

alex-hse-repository

from etna.datasets import TSDataset
from itertools import filterfalse
import pandas as pd
import re


def get_segments(tags: list):
    segments = []
    for tag in tags:
        segments.extend(list(filterfalse(lambda x: tag not in x, data["Page"])))

alex-hse-repository

from etna.analysis import plot_feature_relevance, ModelRelevanceTable
from catboost import CatBoostRegressor

ts = get_ts(segments)
ts.fit_transform(transforms)
plot_feature_relevance(
    ts=ts,
    relevance_table=ModelRelevanceTable(),
    relevance_aggregation_mode="mean",
    relevance_params={

alex-hse-repository

plot_metric_per_segment(metrics, metric_name="SMAPE", top_k=10, ascending=False)

alex-hse-repository

from etna.analysis import plot_metric_per_segment

plot_metric_per_segment(metrics, metric_name="SMAPE", top_k=10, ascending=True)

alex-hse-repository

from etna.models import CatBoostModelMultiSegment
from etna.transforms import SegmentEncoderTransform
from etna.transforms import LagTransform
from etna.analysis import metric_per_segment_distribution_plot
HORIZON = 62

np.random.seed(42)
segments = np.random.choice(data["Page"].values, size=100)
ts = get_ts(segments)
ts.fit_transform(

alex-hse-repository

from etna.analysis import plot_trend
from etna.transforms import LinearTrendTransform, BinsegTrendTransform

ts = get_ts(["List_of_country"])
ts.fit_transform([TimeSeriesImputerTransform(in_column="target", strategy="running_mean", window=3)])
plot_trend(
    ts=ts,
    trend_transform=[
        LinearTrendTransform(in_column="target"),
        BinsegTrendTransform(in_column="target", n_bkps=2, min_size=150),

alex-hse-repository

from etna.models import CatBoostModelPerSegment
from etna.transforms import DateFlagsTransform, FourierTransform

HORIZON = 62
pipeline = Pipeline(
    model=CatBoostModelPerSegment(),
    transforms=[
        DensityOutliersTransform(in_column="target", window_size=30, n_neighbors=9, distance_coef=1),
        TimeSeriesImputerTransform(in_column="target", strategy="running_mean", window=3),
        FourierTransform(period=365.25, order=4),